hexagon logo

Measuring the lowest point in a U shaped groove

Can I vision probe be used to find the lowest point in a groove that is U shaped? or is that better suited for a tactile probe? I need to make sure this u shaped groove is centered between an ID and OD.

Offline with tactile, I can probe a small radius circle, construct a center point, probe the OD, then dimension the center point to the od subtracting the radius. Comes out good offline. Can I do the same thing with the vision probe?

Attached Files
Parents
  • Ok, so my print changed. What the engineer wants to know now is the groove centered in between the ID and OD of this part. C is the ID and D is the OD. At 0.063in from surface B, construct 1 circle on each cone and check the run out of the circles to CD. Pcdmis only lets me choose either C or D and not both.

    This is a ring. Attached is my print showing the cross section and the datums. Surface B isn't really there, its showing in the image because its a cross section. That groove is open. So where datum D and B should intersect, its basically a knife edge. Same goes for where C and B would intersect.

    Because of this, I had to switch to tactile on my optiv to make D a cylinder. When I dimension runout on the constructed circles, it shows absolutely no deviation. I just want to make sure i'm doing anything wrong.

    Plane A is probed, Level Zminus and Z orgin, Cylinder D is probed and XY orgin, nothing to hold rotation.
    Dimension cylinder D, show deviation.
    Probe C as a circle. Dimension, shows deviation.
    Probe 1 cone and then the other.
    Construct a circle on each cone.
    Dimensions runout for each constructed circle and no deviation on runnout.

    Am I missing somthing?

    Attached Files
Reply
  • Ok, so my print changed. What the engineer wants to know now is the groove centered in between the ID and OD of this part. C is the ID and D is the OD. At 0.063in from surface B, construct 1 circle on each cone and check the run out of the circles to CD. Pcdmis only lets me choose either C or D and not both.

    This is a ring. Attached is my print showing the cross section and the datums. Surface B isn't really there, its showing in the image because its a cross section. That groove is open. So where datum D and B should intersect, its basically a knife edge. Same goes for where C and B would intersect.

    Because of this, I had to switch to tactile on my optiv to make D a cylinder. When I dimension runout on the constructed circles, it shows absolutely no deviation. I just want to make sure i'm doing anything wrong.

    Plane A is probed, Level Zminus and Z orgin, Cylinder D is probed and XY orgin, nothing to hold rotation.
    Dimension cylinder D, show deviation.
    Probe C as a circle. Dimension, shows deviation.
    Probe 1 cone and then the other.
    Construct a circle on each cone.
    Dimensions runout for each constructed circle and no deviation on runnout.

    Am I missing somthing?

    Attached Files
Children
No Data